POPLINE Document Number: 018188
Author(s):
McGreevey WP
Source citation:
In: Morse RM, Conniff ML, Wibel J, ed. The urban development of Latin America: 1750-1920. Stanford, California, Stanford Univ. Center for Latin American Studies, 1971. :116-29.
Abstract:
It is contended that certain general patterns of city size distributions can be discovered by statistical analysis of changing absolute and relative population sizes for these cities. The 1st stage of the analysis is to determine the degree to which a system of cities may be said to conform to the Paretian hypothesis of lognormality in their size distribution. It was alleged by Pareto that the lognormal distribution of cities was in fact "normal" for a diversified economy subject to a wide variety of competing forces. He applied his analysis in the late 19th century to Italian cities and found a significant correlation between fact and hypothesis. Concern here is with the opportunity to test the conformance of city size distributions in 8 Latin American countries with the Paretian hypothesis. In the period since World War 2 it has become a matter of conventional wisdom to refer to the hyperencephalization of Latin America's system of cities. The data presented demonstrate that this situation of primacy is a recent phenomenon, i.e., a phenomenon of the 20th century. At the end of the colonial period only 1 country had developed an "excessively" primate pattern. The country is Mexico. The condition of lognormality in the 18th and early 19th centuries is not an idicator that the economies of the other 7 countries were then successfully integrated into a systematic hierarchy. There is some possiblity that such a condition could occur simply by chance, given particularly that the Paretian hypothesis already supposed the prior ranking of cities with regard to size. Yet, the conformance between hypothesis and fact was statistically significant. In the 19th and 20th centuries all city systems stopped being identifiably lognormal and became essentially primate. A hypothesis of growing openness of the economies was considered as a possible cause of the shift from lognormal to primate systems. The higher were per capita exports, the more the city size distribution diverged from the lognormal. The finding conformed to Berry's suggestion that as a system is subjected to the greater influence of a single force, the less it will succumb to general forces of entropy which, he contends, produce the lognormal city distribution. The case of Mexico is anomalous since Mexico was the 1st country in the sample to diverge from the lognormal pattern yet had quite low exports per capita. The data may unfairly understate Mexico's true position as a producer of exports. It may be fruitful to consider other causes for the rise of primacy among Latin American cities.
Keywords:
Statistical StudiesIndex page
Mexico
South America, Central
Latin America
Urbanization
Population Dynamics
Urban Population Distribution
Population Distribution
Commerce
Economic Development
Development Planning
Population Size
Studies
Research Methodology
North America
Americas
Developing Countries
South America
Geographic Factors
Population
Demographic Factors
Macroeconomic Factors
Economic Factors